Aquatic Mammals 2002, 28.3, 221–230

Acomparison of ringed seal ( Phoca hispida)biology on the east and west sides of the North Water Polynya, Baffi nBay

Meike Holst1,2 and Ian Stirling1,3

1 Department of Biological Sciences, University of Alberta, Edmonton, AB, T6G 2E9, Canada 3 Canadian Wildlife Service, 5320 122 Street, Edmonton, AB, T6H 3S5, Canada

Abstract et al., 1994), and their reproductive success and growth can be significantly influenced by changes in In conjunction with the International North Water biological productivity and food abundance (Lowry Polynya Study (NOW) in northern Baffi nBay, we et al., 1980). For example, studies in the Beaufort compared aspects of the biology of ringed seals Sea and Amundsen Gulf have shown that reproduc- ( Phoca hispida)onthe west (Grise Fiord, Nunavut) tive rates sometimes vary markedly in response to and east (Qaanaaq, Greenland) sides of the diff erences in ecological circumstances that appar- polynya. From May to July 1998, Inuit hunters ently a ff ect productivity (e.g., Stirling et al., 1977; collected 99 ringed seals near Grise Fiord, and 100 Smith, 1987; Stirling &Øritsland, 1995; Stirling & ringed seals were collected near Qaanaaq. Adult Lunn, 1997; Kingsley &Byers, 1998). ringed seals from Qaanaaq were significantly In 1998, the International North Water Polynya shorter ( P =0.007) and weighed less ( P =0.0005) Study (NOW) was undertaken to examine the than those from Grise Fiord, but did not diff er in physical and biological mechanisms occurring axillary girth. Female ringed seals from Grise Fiord within the polynya. Several studies have speculated had significantly greater blubber thickness than that the North Water Polynya could be maintained seals from Qaanaaq ( P =0.04). The ovulation rates by sensible heat on the east side and latent heat on for adult female ringed seals were 100% at Grise the west side (e.g., Steff en, 1985; Steff en &Ohmura, Fiord and 93% at Qaanaaq. Evidence of pregnancy 1985; Mysak &Huang, 1992; Darby et al., 1994; in the previous year was found in 80% of females Lewis et al., 1996), which in turn could stimu- from Grise Fiord and in 77% of females from late diff erences in the timing and development of Qaanaaq. The average age of sexual maturity of phytoplankton blooms and food webs within the females at Qaanaaq was earlier than in areas of the polynya. One of the central hypotheses of the NOW Canadian Arctic. Diff erences in the biology of posited that the east side of the polynya was more ringed seals on the west and east sides of the North biologically productive than the west side. If diff er- Water Polynya could reflect variations in biological ences in the level of biological productivity exist productivity within the polynya. within the polynya, they could a ff ect the biology of the ringed seal. Key words: Polynya, ringed seal, Phoca his- The purpose of this study was to compare aspects pida,ovulation rate, sexual maturity, biological of the biology of ringed seals from the east and west productivity. sides of the polynya, including age distribution, growth rate, sexual maturity, condition, ovulation Introduction rate, and pregnancy rate, to examine whether any significant diff erences exist that might relate to the The North Water Polynya in northern Baffi nBay is hypothesized diff erences in productivity. one of the most biologically productive polynyas in the Arctic and an important spring feeding and breeding area for the ringed seal ( Phoca hispida) Materials and Methods (Stirling, 1980, 1997; Stirling et al., 1981). Ringed Comparisons of ringed seal populations were based seals are opportunistic feeders (McLaren, 1958a ; on samples collected on the west side of the North Johnson et al., 1966; Lowry et al., 1978; Weslawski Water Polynya near Grise Fiord, Nunavut 2 Current address: LGL Limited, environmental research (76% 12& N, 83% 06& W) and on the east side near associates, 22 Fisher St., P.O. Box 280, King City, ON, Qaanaaq, Greenland (77% 40& N, 69% 00& W) (Fig.1). L7B 1A6, Canada. From 27 May to 28 June 1998, 99 ringed seals were

 2002 EAAM 222 M. Holst and I. Stirling

Figure 1. Map of the study area and ringed seal collection sites in 1998, adjacent to the North Water Polynya. taken by Inuit hunters near Grise Fiord, of which had to be corrected by adding tail lengths. Tail 95% were collected on landfast ice. Between 8May lengths for each specimen were determined by fit- and 11 July 1998, another 100 ringed seals were ting alinear regression to tail length vs. nose to base collected near Qaanaaq, of which 89% were col- of tail length data from 24 ringed seals taken near lected on landfast ice and 11% were collected in the southwest Baffi nIsland as well as the Belcher water at the ice edge. Hunters were asked to collect Islands (McLaren, unpublished data). The equation seals non-selectively with respect to sex or age class, used was: so that the samples would be representative of the seals available at the time. Samples collected from tail length (cm)=0.0276x+5.0378; R 2 =0.33, each seal included the lower jaw and reproductive tract. All samples were frozen in the field. where xisthe length from nose to base of tail in cm. Measurements for seals collected in Grise Fiord Mean lengths, girths, and weights were calculated followed Scheff er (1967) and Bonner and Laws for adults ( d 6years) only. Mean blubber thickness (1993) and included: standard length, axillary girth, was used as an index of condition and was calcu- and fat thickness (on the ventral mid-line between lated only for animals collected in June, when the front flippers). Seals were not weighed in the animals were collected from both sides of the field, instead weights for specimens from Grise polynya. The condition of pups was assessed separ- Fiord were calculated following methods by ately, since their mean condition at that time McLaren (1958b ): was probably still reflective of the condition of their mothers. Morphometric measurements were log (weight)=3.005 log (length)" 2.9882, compared using Student’s t -tests. To compare growth rates of seals on either side where weight is in pounds and length is in inches. of the polynya, von Bertalanff ygrowth curves, For seals collected in Qaanaaq, axillary girth, fat as parameterized by Hammill et al.(1995), were thickness, and scale weight were recorded. Due to fitted to length vs. age data using the non-linear diff erent measurement techniques employed by col- regression program of SigmaPlot (Ver. 4.0). lectors in Qaanaaq, seal length was measured from Likelihood ratio tests were used to test for statisti- the nose to the base of the tail (instead of to the tip cal diff erences between growth curves and curve of the tail—as measured for standard length), and parameters (Kimura, 1980). Homogeneity of Ringed seal biology in the North Water Polynya 223

Table 1. Comparison between ringed seal populations on the east and west sides of the North Water Polynya (means$ standard deviation; numbers in brackets are sample sizes; M=males, F=females; adults are d 6years old).

Comparison Grise Fiord Qaanaaq Significance*

Sex ratio 43M:56F 55M:45F No Age of adults (yrs) Males 19.4$ 9.4 (27) 15.5$ 10.3 (22) No Females 27.1$ 8.3 (31) 21.4$ 11.8 (14) No Length of adults (cm) Yes Males 136.3$ 14.3 (25) 129.3$ 15.6 (22) Females 136.0$ 13.5 (29) 125.6$ 11.4 (14) Girth of adults (cm) No Males 107.0$ 13.9 (25) 102.5$ 12.7 (21) Females 101.9$ 17.2 (29) 99.2$ 13.7 (13) Weight of adults (kg) Yes Males 75.8$ 24.5 (25) 62.7$ 11.3 (22) Females 75.2$ 22.3 (29) 56.3$ 12.7 (14) Blubber thickness (cm fat) YOY 4.4$ 1.1 (30) 2.4$ 1.5 (7) Yes Males d 1year old 4.3$ 1.2 (24) 4.2$ 0.7 (17) No Females d 1year old 3.7$ 1.2 (34) 4.4$ 1.2 (16) Yes Age of sexual maturity (yrs) N/A 4.70 Ovulation rate 100% (25) 93% (14) No Pregnancy rate 80% (25) 77% (13) No

*Significance of test for Grise Fiord vs. Qaanaaq; P <0.05 was considered significant. variances, an assumption under the Likelihood proportion of females that had acorpus luteum or ratio method (Kimura, 1980), was evaluated using acorpus albicans was used to calculate age-specific an F -test (Sokal &Rohlf, 1981). ovulation rates and the average age of sexual Seal ages were determined using cementum maturity for females (DeMaster, 1978). growth layer groups in canine teeth from the lower jaw (Stewart et al., 1996). Seals less than one year of age are referred to as 0+ or young-of-the-year Results (YOY), seals 1through 5years of age as subadults, and those 6years and older as adults because, on Sex ratio average, that is the age at which females in most Although there were slightly more males than populations reach sexual maturity (e.g., McLaren, females in the sample of ringed seals from Qaanaaq 1958a ;Smith, 1973, 1987). Since sample sizes were and slightly more females than males in the sample not large enough to compare sex-specific age struc- from Grise Fiord, the diff erence was not statistically ture for each individual age class, they were pooled significant in either case (Table 1). Ringed seals as described by Holst et al.(1999), to facilitate from Grise Fiord and Qaanaaq conformed to a1:1 sex ratio (Grise Fiord: " 2 =1.71, df=1, P =0.19; statistical comparisons by age structure: 0+, 1–2, 2 3–5, 6–10, 11–15, 16–20, 21–25, 26–30, 31–35, 36– Qaanaaq: " =1.00, df=1, P =0.32). 40, and 41+years of age. Age distributions were tested using aLikelihood chi-square test. Ages of Age distribution males and females at the two locations were com- The age distributions for males and females pared with Mann-Whitney U -tests, and sex ratios (Fig. 2) diff ered significantly between the two were tested using Pearson’s chi-square test. locations (males: " 2 =22.55, df=10, P =0.01; Ovaries were sectioned at 2-mm intervals and females: " 2=46.63, df=10, P <0.0001). In the Grise were examined macroscopically for the presence of Fiord sample, 32% were YOY and 9% were sub- corpora lutea (indicative of ovulation in the current adults, whereas the proportion of YOY in the year) and corpora albicantia (indicative of preg- Qaanaaq sample was 7% and 56% were subadults. nancy in the previous year) (McLaren, 1958a ). The There was agreater proportion (18%) of older seals 224 M. Holst and I. Stirling

Figure 2. Age distributions of (a) male (n=41) and female (n=56) ringed seals collected in Grise Fiord and (b) male (n=55) and female (n=43) ringed seals collected in Qaanaaq during May–July, 1998.

( d 31 years) in the Grise Fiord sample than in the with those from Qaanaaq ( t =2.77, df=88, Qaanaaq sample (6%). P=0.007). Although the axillary girths of seals On average, adult females collected near Grise collected in Grise Fiord were slightly greater, Fiord were significantly older than adult males there were no significant diff erences between the ( P =0.005; Table 1). Adult females collected near mean girths of seals collected in Qaanaaq and Qaanaaq also were on average older than adult those collected in Grise Fiord ( t=0.93, df=86, males, but the diff erence was not statistically signifi- P =0.36). cant ( P=0.14). The mean ages of adults of the same The von Bertalanff ygrowth curves diff ered sig- sex were not significantly diff erent between the two nificantly between ringed seals from the two sides of locations ( P =0.14). The maximum ages for males the polynya (Fig. 3). Ringed seals from Qaanaaq and females collected near Grise Fiord were 34 and had asignificantly higher growth rate ( P =0.006) 43, respectively, and for Qaanaaq, the maximum than seals from Grise Fiord (Table 2). However, ages were 40 and 38 years, respectively. seals from Grise Fiord were significantly longer at birth ( P =0.007) and reached asignificantly greater Morphometrics asymptotic length ( P =0.002) than seals from Adult male and female ringed seals did not diff er Qaanaaq (Table 2). significantly in mean length or axillary girth at either Grise Fiord or Qaanaaq (Table 1), so the Body condition data were pooled for each location. Seals from The mean weights of adult males and females did Grise Fiord were significantly longer compared not diff er significantly at either location (Table 1), Ringed seal biology in the North Water Polynya 225

Figure 3. Von Bertalanff ygrowth curves for ringed seals collected near Grise Fiord and Qaanaaq.

Table 2. Parameter estimates ( $ SE) and R 2 values for the von Bertalanff ygrowth curves for ringed seals collected around Qaanaaq and Grise Fiord; L ` =asymptotic length (cm), k o =absolute growth rate at birth (cm/year), and l o =length at birth (cm).

2 Sample Sample size L ` k o l o R

Grise Fiord 89 137.50$ 1.99 9.60$ 2.43 94.53$ 2.28 0.72 Qaanaaq 98 128.22$ 1.95 26.73$ 6.09 81.50$ 4.13 0.58

thus the data were pooled. The mean calculated Reproduction weight of adult seals from Grise Fiord was signifi- The average age of sexual maturity for female cantly higher than the mean scaled weight of adults ringed seals from Grise Fiord could not be deter- from Qaanaaq ( t -test=3.64, df=88, P =0.0005). mined, since no females 5–8 years old were There were no significant diff erences in the mean sampled. Ovulation had not taken place in the only blubber thickness between males and females d 1 4-year-old female but had occurred in the single year old from the same location (Table 1). The 9-year-old. All females d 9years old had ovulated, mean blubber thickness of males d 1year did not and acorpus albicans was found in the ovaries of diff er between the two locations ( t =0.33, df=39, 80% of them (Table 1). The average age of sexual P =0.74). The mean blubber thickness of females maturity for female ringed seals from Qaanaaq was d 1year old from Grise Fiord was significantly 4.7 years. Ovulation (Table 1) had occurred in lower than from Qaanaaq ( t =2.14, df=48, 29% of 4-year-old females, all of the 5-year-olds, P =0.04). YOY from Grise Fiord had agreater and 93% of females d 6years old. Acorpus albicans mean blubber thickness than those from Qaanaaq was found in the ovaries of 77% females d 6years ( t =4.03, df=35, P =0.0003; Table 1). old. 226 M. Holst and I. Stirling

Discussion diff erences in the habitat in which the two samples were taken. There are likely to be fewer subadults The sex ratios of ringed seals from Grise Fiord and present in habitat sampled farther from the floe Qaanaaq did not diff er from unity, as is character- edge, such as near Grise Fiord, than in habitats that istic of ringed seal populations throughout the are sampled closer to the floe edge, as near Canadian Arctic (Smith, 1973, 1987; Holst et al., Qaanaaq. 1999), as well as other locations (Helle, 1974; Adult male and female ringed seals collected at Fedoseev, 1965; Lydersen &Gjertz, 1987). Grise Fiord and Qaanaaq were similar in length The mean age of adult female ringed seals col- and weight. Smith (1973) reported that in Davis lected near Grise Fiord was significantly older than Strait, males were significantly larger than females, for adult males, as has also been found in studies of and Goodyear (1999) showed that males from ringed seals from other areas (Helle, 1974, 1979; Baffin Bay reached agreater asymptotic length than Lydersen &Gjertz, 1987). Ringed seals are slightly females. At Point Hope, Alaska, Johnson et al. polygynous (Stirling, 1983), and male seals main- (1966) found that males were heavier than females tain and defend underwater territories around from November to May, but females were heavier breathing holes (Stirling, 1973, 1977). Thus, males than males in June. In Svalbard, females collected might be expected to have shorter longevity than between March and July were significantly heavier females. than males (Lydersen &Gjertz, 1987). Females Alow proportion of YOY in the samples of could be heavier than males during the spring ringed seals from Qaanaaq and Grise Fiord was because of diff erent behaviour associated with expected, since YOY are generally less available for mating and pupping (Lydersen &Gjertz, 1987). sampling during late spring. Some pups use sub- Ringed seals show considerable variation in adult nivean lairs until about mid May, where they are length among populations throughout their cir- less visible and thus less vulnerable to hunters. After cumpolar range (e.g., McLaren, 1958a ,1993; that time, most pups are weaned and molted, and Fedoseev, 1965; Fedoseev &Nazerenko, 1970; spend more time in the water feeding (Smith, 1973). Smith, 1973, 1987; Helle, 1979; Finley et al., 1983; Thus, the low proportion of YOY in the spring Frost &Lowry, 1981; Lydersen &Gjertz, 1987; sample from Qaanaaq is consistent with results Goodyear, 1999). Ringed seals from Grise Fiord obtained in other studies (e.g., Smith, 1973; were similar in size to other seal populations from Lydersen &Gjertz, 1987). the Canadian northeastern Arctic (McLaren, However, the proportion of YOY in the Grise Fiord 1958a ,1993; Finley et al., 1983; Goodyear, 1999). sample was higher than expected. This discrepancy However, adult ringed seals from Qaanaaq were could be at least partially explained by diff erences in shorter than those from Grise Fiord, but were the landfast ice habitat, where most of the sampling comparable in size to those from Svalbard took place in each area. At Qaanaaq, seals were taken (Lydersen &Gjertz, 1987), the Beaufort Sea mainly in the breeding habitat in the landfast ice, and Amundsen Gulf (Smith, 1987), southeastern although some were taken further out toward the Baffi nBay (McLaren, 1958a ,1993; Smith, 1973; unstable floe edge, where pupping is expected to occur Goodyear, 1999), and the pack ice of Baffi nBay less frequently. In comparison, at Grise Fiord, seals (Finley et al., 1983). Correspondingly, the growth were taken further from the floe edge in Jones Sound curves suggested that pups sampled at Qaanaaq and more in fiords on southern Ellesmere Island, were smaller than those at Grise Fiord. where pupping was likely more prevalent, thus making The growth curve parameters may have been pups more vulnerable to sampling. influenced by the age composition of the samples. The Grise Fiord sample had asignificantly lower For example, the lack of newborn pups and prepon- proportion (9%) of subadults than did the sample derance of weaned pups in the sample are respon- from Qaanaaq, which is typical for spring samples sible for the parameter l o (length at birth) being taken on landfast ice (e.g., McLaren, 1958a ;Smith, more representative of length of weaned rather than 1973, 1987; Smith &Stirling, 1975). Under- newborn pups. However, asymptotic length tends sampling of subadults in Grise Fiord could have to be estimated more robustly (McLaren 1993). occurred because during spring, subadults tend to Diff erences in size among populations of ringed be excluded from the breeding areas in landfast ice seals may be due to several factors, including gen- to unstable moving ice areas further o ff shore where etic variation, stability of breeding habitat, density- they are encountered less frequently by hunters dependent e ff ects resulting from diff erential levels (McLaren, 1958a ;Smith, 1973, 1987; Smith & of harvesting, and diff erences in food availability. Stirling, 1975). As with the diff erence in proportions Finley et al.(1983) speculated that diff erences in of pups in the samples from the two areas, the size between landfast ice and pack ice seals could be diff erence in the proportions of subadults is also due to genetic diff erences, but demonstrated none. likely related to relatively small, but important Preliminary analyses of samples collected during Ringed seal biology in the North Water Polynya 227 this study indicate that ringed seals from Qaanaaq In accordance with the NOW hypothesis, if bio- and Grise Fiord are genetically similar (Davis et al., logical productivity is higher on the east side of the 1999), suggesting that at least some interbreeding polynya, food biomass and availability could be occurs between ringed seals from either side of the greater on that side. Lewis et al.(1996) demon- polynya. Therefore, any existing diff erences in size strated an increase in phytoplankton biomass on the are likely the result of environmental factors. east side of the North Water Polynya in May. An McLaren (1958a )and Pastukhov (1969) sug- early bloom on the east side could lead to greater gested that diff erences in adult ringed seal sizes food availability early in the spring and an increase occur because of variation in the stability of ice in fish biomass. Consequently, ringed seals sampled conditions in the pupping habitat, i.e., pups born in near Qaanaaq could have access to amore abun- more stable ice areas were likely to be nursed longer dant food supply and have faster early growth rates. than those in less stable habitat and thus would be Faster growth rates also could be due to increased larger at weaning. In general, coastal areas are more consumption of prey with higher energy con- stable than drifting pack ice and thus are thought to tent, such as Arctic cod ( Boreogadus saida)(e.g., be better habitat for pupping (McLaren, 1958a ). Bradstreet &Finley, 1983). Diff erences in the diet of Finley et al.(1983) and Fedoseev (1975) found seals ringed seals on either side of the polynya have from drift or pack ice to be smaller than seals from already been demonstrated. Ringed seals near coastal landfast ice, and Smith (1973) concluded Qaanaaq had agreater proportion of Arctic cod in that diff erences in size between adult seals from their diet compared to seals near Grise Fiord (Holst southeastern Baffi nIsland were due to diff erences in et al., 2001). ice stability. Ice conditions could partially explain Ringed seals from Svalbard, ahighly productive size diff erences between ringed seals from the east area, also have high growth rates (Lydersen & and the west sides of the polynya. Although ringed Gjertz, 1987), as do seals from Arctic Bay, seals collected in both Qaanaaq and Grise Fiord although data on primary productivity are not were harvested in generally similar landfast ice available for that area (Goodyear, 1999). However, areas, the landfast ice break-up in the Qaanaaq even though ringed seal populations from Svalbard, collection area occurred on approximately 26 May Arctic Bay, and Qaanaaq had higher growth rates 1998, whereas break-up in the Grise Fiord collec- for young seals than those from other areas, they tion area occurred around 4June 1998. Therefore, if did not attain higher asymptotic lengths when com- early break-up in the Qaanaaq area led to ashorter pared to slower growing populations (see review by suckling period, then pups would have been weaned McLaren, 1993). It is not apparent why ringed seal earlier in that area. However, it remains uncertain populations with high growth rates do not reach whether ice conditions would cause the significant greater asymptotic lengths. diff erence in seal size observed, or whether other Age of sexual maturity varies among and within factors also a ff ected seal growth in the area. species of pinnipeds. Factors proposed to a ff ect the Goodyear (1999) speculated that variation in onset of sexual maturity are density-dependence stability of sea ice pupping habitat did not and overall ecosystem productivity. Age of sexual adequately explain diff erences in seal lengths for maturity of seals has been linked to adensity- populations in the northeastern Canadian Arctic. dependent response, decreasing during periods He proposed that predation pressure from polar of low population size (e.g., Laws, 1953, 1956a ; bears ( Ursus maritimus)and Inuit hunters, and Carrick et al., 1962; Capstick &Ronald, 1982; Sjare availability of food resources, have agreater e ff ect et al., 1996). In female ringed seals of diff erent on the growth of ringed seals than ice conditions. geographic regions, the average age of sexual Although it is possible that human hunting pressure maturity varies greatly (Table 3), ranging from 4.4 may be higher in the Qaanaaq area than in Grise years in Svalbard (Lydersen &Gjertz,1987) to 7.0 Fiord (e.g., Teilmann &Kapel, 1998; Finley & years in Cumberland Sound, Baffi nBay (Smith, Miller, 1980; Stewart et. al.,1986), that of polar 1973). Growth rates and sizes of seals from Grise bears may be lower, possibly cancelling each other Fiord are similar to those from surrounding areas out (e.g., Stirling &Øritsland, 1995; Taylor et al., in the Canadian northeastern Arctic (McLaren, 2001). Regardless, it is unlikely that human or bear 1958a ,1993), suggesting that average age of sexual predation explain the observed diff erences. Conse- maturity is likely similar as well. The relatively quently, other factors such as food availability and higher early growth rates for ringed seals from the quality of the prey are more likely to determine Qaanaaq are consistent with achieving sexual growth in these populations. For marine mammals maturity at an earlier age than those from Grise in general, nutrition is thought to be the most Fiord, as would be predicted by Laws (1956b ). important factor a ff ecting the shape of growth Ringed seals from Qaanaaq reached sexual curves, with better nutrition leading to more rapid maturity earlier than those from other areas in the growth rates (Laws, 1956b ). Arctic, except for Svalbard (Lydersen &Gjertz, 228 M. Holst and I. Stirling

Table 3. Age-specific ovulation rates of female ringed seals from selected areas of the Arctic and average age of sexual maturity.

%Ovulation in age class Average age Data source of sexual Location 34 5678910 maturity

Northwestern Hudson Bay 17 29 33 57 100 100 100 100 5.6 Holst et al., 1999 Foxe Basin, Hudson Strait 0122278100 100 100 100 5.9 McLaren, 1958 Cumberland Sound, Baffi n 13 20 12 58 92 100 100 7.1 Smith, 1973 Bay Hoare Bay, Davis Strait 40 33 80 100 75 100 100 5.7 Smith, 1973 Ellesmere Island and Northwest 22 53 72 81 98 ca. 5.7 Miller &Finley, Baffi nIsland 1982 Grise Fiord, Southern Ellesmere 00 100 This study Island Qaanaaq, Northwestern 029100 100 100 100 100 4.7 This study Greenland Point Hope, Alaska 08605593100 100 100 5.8 Johnson et al., 1966 Amundsen Gulf and Southeastern 0255361100 100 100 100 5.6 Smith, 1987 Beaufort Sea Svalbard 20 60 80 100 100 100 100 100 4.4 Lydersen &Gjertz, 1987

1987). It is unlikely that density-dependent factors from both locations were likely in good condition, reduced the age of sexual maturity at Qaanaaq, since both had high ovulation and pregnancy rates. since ringed seals are thought to be numerous in The mean condition of YOY from Grise Fiord was that area and hunting pressures are low. higher than that for YOY collected in Qaanaaq, Laws (1956b )suggested that nutrition was likely which is consistent with the larger size of the adult the most important factor that caused early sexual females from Grise Fiord. YOY sampled in maturation, by increasing growth rates, whereas a Qaanaaq could have had less blubber (or been in decrease in food availability would lead to delayed poorer condition) than pups in Grise Fiord, if they sexual maturation (Eberhardt, 1977). Lydersen & were indeed weaned earlier, and had been foraging Gjertz (1987) suggested that early sexual matura- independently for some time. tion of ringed seals in Svalbard could be explained by the abundance of food in that area. Thus, in Qaanaaq, the hypothesized increase in food avail- Conclusions ability on the east side of the polynya could explain early sexual maturation in ringed seals. The major diff erences between ringed seals inhabit- The mean weights for seals collected in Grise ing the east and west sides of the North Water Fiord were similar to those of ringed seals collected Polynya were in growth rates, average age of sexual at asimilar time of year near Pond Inlet, northern maturity, and total body lengths. Seals on the east Baffi nIsland (Finley et al., 1983) and those of side of the polynya had faster early growth rates Qaanaaq are comparable to mean weights of seals and appeared to attain sexual maturity earlier, but sampled at Svalbard (Lydersen &Gjertz, 1987). had shorter adult lengths than seals on the west side The mean blubber thicknesses of ringed seals col- of the polynya. Ice stability and food availability lected near both Grise Fiord and Qaanaaq were likely had the greatest e ff ect on growth rates. similar to the blubber thickness of seals collected Ringed seals on both sides of the polynya in the same season in the Barents Sea (Wathne, appeared to be in similar body condition and 1997) and southeastern Baffi nIsland (Goodyear, had high ovulation and pregnancy rates. Since 1999). polynyas are considered to be more productive The reproductive status of female ringed seals has than surrounding non-polynya areas, an increase in been shown to be a ff ected by their condition, with a biological productivity within the polynya would decrease in reproductive rates when body condition likely have alimited e ff ect on the biology of ringed is low (e.g., Smith, 1987; Kingsley &Byers, 1998). seals on either the east or west sides of the North Even though adult females from Grise Fiord had Water Polynya, since an adequate food supply is less blubber than those from Qaanaaq, females most likely available throughout the entire area. Ringed seal biology in the North Water Polynya 229

Acknowledgments Fisheries Research Board of Canada Translation Series No. 2411 (1973). We thank the Hunters and Trappers Organization Finley, K. J. &Miller, G. W. (1980) Wildlife Harvest in Grise Fiord and R. Dietz and C. Sonne-Hansen Statistics from Clyde River, Grise Fiord and Pond for collecting specimens and measurements in Grise Inlet, 1979. Report by LGL Ltd., Toronto. Eastern Fiord and Qaanaaq, respectively. We also thank Arctic Marine Environmental Studies, EAMES North I. McLaren for providing us with several seal Reports, EN 15, Wildlife Harvest Statistics. measurements. W. Calvert, M. McPherson, and Finley, K. J., Miller, G. W., Davis, R. A. &Koski, W. R. C. Spencer helped with age determination. S. Barry, (1983) Adistinctive large breeding population of ringed V. Moulton, A. Serrano, and M. Goodyear pro- seals ( Phoca hispida)inhabiting the Baffi nBay Pack ice. vided statistical advice. J. Thomas and L. Lowry, as Arctic 36, 162– 173. Goodyear, M. A. (1999) Variation in growth and seasonal well as an anonymous referee, provided construc- condition of ringed seal, Phoca hispida,from the tive criticisms of the manuscript. This project was Canadian Arctic. MSc Thesis, University of Manitoba, supported by the Canadian Wildlife Service and Winnipeg, MB. 111 pp. the Natural Sciences and Engineering Research Hammill, M. O., Kingsley, M. C. S., Beck, G. G. &Smith, Council of Canada. T. G. (1995) Growth and condition in the Northwest Atlantic harp seal. Canadian Journal of Fisheries and Aquatic Sciences 52, 478–488. Literature Cited Helle, E. (1974) On the biology of the ringed seal Pusa Bonner, W. N. &Laws, R. M. (1993) Morphometrics, hispida in the Bothnian Bay. Proceedings from the specimen collection and preservation. In: R. M. symposium on the seal in the Baltic. National Swedish Laws (ed). Antarctic seals,pp. 161–171. Cambridge Environment Protection Board PM 591: 38–42. University Press, Cambridge. Helle, E. (1979) Growth and size of the ringed seal Phoca Bradstreet, M. S. W. &Finley, K. J. (1983) Diet of ringed (Pusa) hispida Schreber in the Bothnian Bay, Baltic. seals ( Phoca hispida)inthe Canadian High Arctic. Zeitschrift für Säugetierkunde 44, 208–220. Report by LGL Ltd., Toronto, Ontario. Holst, M., Stirling, I. &Calvert, W. (1999) Age structure Capstick, C. K. &Ronald, K. (1982) Modelling seal and reproductive rates of ringed seals ( Phoca hispida) populations for herd management. FAO Fisheries on the northwestern coast of Hudson Bay in 1991 and Series 4, 177–190. 1992. Marine Mammal Science 15, 1357–1364. Carrick, R., Csordas, S. E. &Ingham, S. E. (1962) Studies Holst, M., Stirling, J. &Hobson, K. A. (2001) Diet of on the southern elephant seal, Mirounga leonina ringed seals ( Phoca hispida)onthe east and west sides (L), Part 4: Breeding and development. Commonwealth of the North Water Polynya, northern Baffi nBay. scientific and industrial research organization Marine Mammal Science 17,888–908. (C.S.I.R.O.). Wildlife Research 7, 161–197. Johnson, M. L., Fiscus, C. H., Ostenson, B. T. &Barbour, Darby, M. S., Willmott, A. J. &Mysak, L. A. (1994) A M. L. (1966) Marine Mammals. In: N. J. Wilimovsky & non-linear steady-state model of the North Water J. N. Wolfe (eds.) Environment of the Cape Thompson Polynya, Baffi nBay. Journal of Physical Oceanography region, Alaska U.S.,pp. 877–924. Atomic Energy 24, 1011–1020. Commission, Oak Ridge, TN. Davis, C., Stirling, I. &Strobeck, C. (1999) Genetic Kimura, D. K. (1980) Likelihood methods for the von diversity and population structure of two arctic ice- Bertalanff ygrowth curve. Fishery Bulletin 77, 765–776. breeding phocids: bearded seals ( Erignathus barbatus) Kingsley, M. C. S. &Byers, T. J. (1998) Failure and ringed seals ( Phoca hispida). Abstract for the of reproduction in ringed seals ( Phoca hispida)in 13th Biennial Conference on the Biology of Marine Amundsen Gulf, Northwest Territories in 1984–1987. Mammals, Nov. 28–Dec. 3, 1999, Maui, Hawaii. In: M. P. Heide-Jørgensen &C.Lydersen (eds.) Ringed DeMaster, D. P. (1978) Calculation on the average age of seals (Phoca hispida) in the North Atlantic,pp. 197– sexual maturity in marine mammals. Journal of the 210. NAMMCO Scientific Publications Volume 1. The Fisheries Research Board of Canada 25, 912–915. North Atlantic Marine Mammal Commission, Tromsø. Eberhardt, L. L. (1977) Optimal policies for conservation Laws, R. M. (1953) The elephant seal ( Mirounga leonina of large mammals, with special reference to marine Linn.) Part 1: Growth and age. Falkland Islands ecosystems. Environmental Conservation 4, 205–212. Dependencies Survey, London. Scientific Report No. 8. Fedoseev, G. A. (1965) Age and sex composition of the kill of the ringed seal ( Phoca hispida ochotensis Pall.) Laws, R. M. (1956a )The elephant seal ( Mirounga leonina as an index of the age structure of the population. Linn.). Falkland Islands Dependencies Survey, Fisheries Research Board of Canada Translation Series London. Scientific Report No. 2. No. 799 (1966). Laws, R. M. (1956b )Growth and sexual maturity in Fedoseev, G. A. (1975) Ecotypes of the ringed seal ( Pusa aquatic mammals. Nature 178, 193–194. hispida Schreber, 1777) and their reproductive capabili- Lewis, E. L, Ponton, D., Legendre, L. &LeBlanc, ties. In: K. Ronald &A.W.Mansfield (eds.) Biology of B. (1996) Springtime sensible heat, nutrients and the Seal,pp. 156–160. Rapports et Proces-verbaux des phytoplankton in the Northwater Polynya, Canadian Réunions, Conseil International pour L’Exploration de Arctic. Continental Shelf Research 16, 1775–1792. la Mer 169. Lowry, L. F., Frost, K. J. &Burns, J. J. (1978) Food of Fedoseev, G. A. &Nazarenko, Y. I. (1970) On the ringed seals and bowhead whales near Point Barrow, intraspecific structure of ringed seal in the Arctic. Alaska. Canadian Field-Naturalist 92, 67–70. 230 M. Holst and I. Stirling

Lowry, L. F., Frost, K. J. &Burns, J. J. (1980) Variability Stirling, I. (1977) Adaptations of Weddell and ringed seals in the diet of ringed seals, ( Phoca hispida), in Alaska. to exploit the polar fast ice habitat in the absence or Canadian Journal of Fisheries and Aquatic Sciences 37, presence of surface predators. In: G. A. Llano (ed.). 2254–2261. Adaptations within Antarctic ecosystems,pp. 741–748. Lydersen, C. &Gjertz, I. (1987) Population parameters of Proceedings of the 3rd Scientific Committee on ringed seals ( Phoca hispida Schreber, 1775) in the Antarctic Research symposium on Antarctic Biology, Svalbard area. Canadian Journal of Zoology 65, 1021– Washington, DC, August 26–30, 1974. 1027. Stirling, I. (1980) The biological importance of polynyas McLaren, I. A. (1958a )The biology of the ringed seal in the Canadian Arctic. Arctic 33, 303–315. ( Phoca hispida Schreber) in the eastern Canadian Stirling, I. (1983) The evolution of mating systems in Arctic. Bulletin of the Fisheries Research Board of pinnipeds. In: J. F. Eisenberg &D.G.Kleiman (eds.) Canada No. 118. Recent advances in the study of mammalian behavior, McLaren, I. A. (1958b )The economics of seals in the pp. 489–527. Special Publication, American Society of eastern Canadian arctic. Fisheries Research Board of Mammalogy No. 7. Canada Circular No. 1. Stirling, I. (1997) The importance of polynyas, ice edges, McLaren, I. A. (1993) Growth in pinnipeds. Biological and leads to marine mammals and birds. Journal of Review 6, 1–79. Marine Systems 10, 9–21. Mysak, L. A. &Huang, F. (1992) Alatent- and sensible- Stirling, I., &Lunn, N. J. (1997) Environmental fluctua- heat polynya model for the North Water, northern Baffin tions in arctic marine ecosystems as reflected by varia- Bay. Journal of Physical Oceanography 22, 596–608. bility in reproduction of polar bears and ringed seals. Pastukhov, V. D. (1969) Onset of sexual maturity of In: S. J. Woodin &M.Marquiss (eds.) Ecology of female ringed seals. Fisheries Research Board of Arctic Environments,pp. 167–181. Blackwell Science Canada Translation Series No. 1474 (1970). Ltd., Oxford. Scheff er, V. B. (1967) Standard measurements of seals. Stirling, I. &Øritsland, N. A. (1995) Relationships Journal of Mammalogy 48, 459–462. between estimates of ringed seal ( Phoca hispida) Sjare, B., Stenson, G. B. &Warren, W. G. (1996) and polar bear ( Ursus maritimus)populations in the Summary of female harp seal reproductive parameters Canadian Arctic. Canadian Journal of Fisheries and in the Northwest Atlantic. NAFO Scientific Council Aquatic Sciences 52, 2594–2612. Studies 26, 41–46. Stirling, I., Archibald, R. &DeMaster, D. (1977) Smith, T. G. (1973) Population dynamics of the ringed Distribution and abundance of seals in the eastern seal in the Canadian eastern Arctic. Fisheries Research Beaufort Sea. Journal of the Fisheries Research Board of Board of Canada Bulletin No. 181. Canada 34, 976–988. Smith, T. G. (1987) The ringed seal, Phoca hispida,of Stirling, I., Cleator, H. &Smith, T. G. (1981) Marine the Canadian western Arctic. Canadian Bulletin of mammals. In: I. Stirling &H.Cleator (eds.) Fisheries and Aquatic Sciences No. 216. Polynyas in ,pp. 45–58. Canadian Wildlife Smith, T. G. &Stirling, I. (1975) The breeding habitat of the Canadian Arctic Service Occasional Paper No. 45. the ringed seal ( Phoca hispida). The birth lair and associated structures. Canadian Journal of Zoology 53, Taylor, M. K., Akeeagok, S., Andriashek, D., Barbour, 1297–1305. W., Born, E. W., Calvert, W., Cluff ,H.D., Ferguson, Sokal, R. R. &Rohlf, F. J. (1981) Biometry: the principles S., Laake, J., Rosing-Asvid, A., Stirling, I. &Messier, and practice of statistics in biological research. W.H. F. (2001) Delineation of Canadian and Greenland Freeman and Company, San Francisco. polar bear ( Ursus maritimus)populations using cluster Steff en, K. (1985) Warm water cells in the North analysis of movements. Canadian Journal of Zoology Water, northern Baffi nBay during winter. Journal of 79, 690–709. Geophysical Research 90, 9129–9136. Teilmann, J. &Kapel, F. O. (1998) Exploitation of Steff en, K. &Ohmura, A. (1985) Heat exchange and ringed seals ( Phoca hispida)inGreenland. In: M. P. surface conditions in North Water, northern Baffi n Heide-Jørgensen &C.Lydersen (eds.) Ringed seals in Bay. Annals of Glaciology 6, 178–181. the North Atlantic,pp. 130–151. NAMMCO Scientific Stewart, R. E. A., Richard, P., Kingsley, M. C. S. & Publications Volume 1. The North Atlantic Marine Houston, J. J. (1986) Seals and sealing in Canada’s Mammal Commission, Tromsø. northern and arctic regions. Canadian Technical Wathne, J. A. (1997) Feeding habits, condition and poss- Report of Fisheries and Aquatic Sciences 1463:1–31. ible resource partitioning of ringed seals Phoca hispida Stewart, R. E. A., Stewart, B. E., Stirling, I. &Street, E. in drift ice areas in the Barents Sea. Cand. scient thesis. (1996) Counts of growth layer groups in cementum and Institute of Aquatic Resources and Environmental dentine in ringed seals ( Phoca hispida). Marine Mammal Biology, the Norwegian College of Fishery Science, Science 12, 383–401. University of Tromsø, Tromsø. 59 pp. Stirling, I. (1973) Vocalizations in the ringed seal ( Phoca Weslawski, J. M., Ryg, M., Smith, T. G. &Øritsland, hispida). Journal of the Fisheries Research Board of N. A. (1994) Diet of ringed seals ( Phoca hispida)ina Canada 30, 1592–1594. Fjord of West Svalbard. Arctic 47, 109–114.